CN105913093A - Template matching method for character recognizing and processing - Google Patents
Template matching method for character recognizing and processing Download PDFInfo
- Publication number
- CN105913093A CN105913093A CN201610284102.XA CN201610284102A CN105913093A CN 105913093 A CN105913093 A CN 105913093A CN 201610284102 A CN201610284102 A CN 201610284102A CN 105913093 A CN105913093 A CN 105913093A
- Authority
- CN
- China
- Prior art keywords
- image
- template
- identified
- current
- characteristic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/24—Aligning, centring, orientation detection or correction of the image
- G06V10/243—Aligning, centring, orientation detection or correction of the image by compensating for image skew or non-uniform image deformations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The present invention discloses a template matching method for character recognizing and processing, belonging to the technical field of pattern recognition and image processing. The method comprises the following steps: preprocessing the characteristics extraction of an inputted template and a to-be-recognized image; performing rectangle frame calibration by using adjacent communicable regions in the image and word region as units, and obtaining a plurality of characteristic rectangles; and then performing multiple times of normalizations of coordinate origins and scales of the current template image and the image to be recognized; calculating the similarity of the characteristic set for each normalization; and taking the largest similarity of characteristic set as the matching similarity of the current template image; and finally obtaining the template with the maximum matching similarity of all the template images that is taken as the best matching template of the current image to be recognized. The character recognizing module of the invention is used for character recognizing and processing, especially for a medical examination sheet recognition system, which can overcome the influences of factors such as scene image wrinkling, scale change and light change on character recognition, thus improving the character recognition rate.
Description
Technical field
The invention belongs to pattern recognition and technical field of image processing, the template matching being specifically related in character recognition technology
Technology.
Background technology
Along with the high speed development of the progress of information technology, particularly artificial intelligence technology, OCR (optical character recognition) technology
Having been achieved for a lot of achievement in information identification field, also step into the practical stage of socialization simultaneously, OCR technique is successfully applied to
In society's all trades and professions, also there is universal application in medical industry.
From the viewpoint of statistical-simulation spectrometry, the problem identifying an actually pattern classification of character.Both at home and abroad
Scholar proposes many different identifying schemes, is broadly divided into recognition methods based on grader and side based on template matching
Method.
The method utilizing grader to be identified can obtain preferable recognition result, but the method for grader needs big
The learning sample of amount is trained.Template matching algorithm is typically to mate bianry image, is to realize discrete input pattern
One of effective way of classification, essence is certain similarity between tolerance input pattern and sample, and taking similarity the maximum is
Input pattern generic.It, according to the visual pattern extraction feature of character, is identified by relevant matches principle.
Angularly considering from real-time, algorithm complexity, template matching algorithm disclosure satisfy that the task of character recognition.But
Template matching also has its defect, i.e. quick in yardstick polytropy and the illumination polytropy of different scene images to same class character
Sense, easily produces coupling deviation.
Summary of the invention
The technical problem to be solved is to provide a kind of template matching method processed for Text region, the party
Method carries out mating optimizing, to determine the Optimum Matching template of image to be identified (test sample) in template set.
The template matching method processed for Text region of the present invention, comprises the following steps:
Image to be identified and several template images (quantity of template image is configured based on practical application request) are entered
Row Image semantic classification, described Image semantic classification includes image slant correction, image denoising, image gray processing, image binaryzation;
Image to be identified and each template image are carried out feature extraction, obtain characteristic of correspondence set:
To the character area of each template image in units of adjacent connected domain, carry out the demarcation of rectangle frame, obtain template
The characteristic set of imageWhereinRepresent that the lateral coordinates of jth feature square in current template image T, longitudinal coordinate, pixels across are long
Degree, longitudinal length in pixels, j={1,2,3 ..., n}, n represent the number of characteristic rectangle in template image T;
To the character area of image to be identified in units of adjacent connected domain, carry out the demarcation of rectangle frame, obtain waiting to know
The characteristic set of other imageWhereinRepresent the lateral coordinates of ith feature rectangle, longitudinal coordinate, horizontal picture in current image S to be identified
Element length, longitudinal length in pixels, i={1,2,3 ..., m}, m represent the number of characteristic rectangle in image S to be identified;
Template image and image to be identified carry out repeatedly zero normalized, dimension normalization processes, and often enters
Normalized of row, then calculate a characteristic set similarity, takes maximum characteristic set similarity as current template figure
The matching similarity of picture;Take the template image corresponding to maximum match similarity and correctly mate mould as current image to be identified
Plate;Wherein, the computing formula of characteristic set similarity sim is:αi∩βjRepresent current figure to be identified
The rectangular area overlapping area size of the ith feature rectangle of picture and the jth characteristic rectangle of current template image, αi∪βiTable
Show the rectangular area union area of the ith feature rectangle of front image to be identified and the jth characteristic rectangle of current template image
Size.
Further, in order to reduce computation complexity, when dimension normalization processes, first scaling is set to R also
Carry out a dimension normalization to process, then obtain optimal scaling based on climbing algorithm screening and carry out the yardstick normalizing of correspondence
Change processes, wherein R=ws/wT, parameter wsFor the picture traverse of current image to be identified, wTHorizontal image for current template image
Width;Or R=hs/hT, parameter hsFor the picture traverse of current image to be identified, hTHorizontal figure image width for current template image
Degree.
In sum, owing to have employed technique scheme, the invention has the beneficial effects as follows: solve the pleat of scene image
The impact that Text region is caused by the factors such as wrinkle, yardstick is changeable and illumination is changeable, is effectively improved character identification rate.
Accompanying drawing explanation
Fig. 1 be the specific embodiment of the invention realize block diagram.
Fig. 2 be the present invention a kind of normalization anchor point and Normalized Scale choose coordinate diagram.
Detailed description of the invention
For making the object, technical solutions and advantages of the present invention clearer, below in conjunction with embodiment and accompanying drawing, to this
Bright it is described in further detail.
The template matching method of the present invention being used for Text region process, it is as follows that it implements step:
S1: Image semantic classification process:
S1-1: input image to be identified.Image the most to be identified can be by online picture pick-up device collection, it is also possible to be
The scene image preserved.In present embodiment, the image pattern of collection utilizes camera to shoot, and image pattern is from 13 kinds
Different types of papery medicalization verification certificate.For same category of same laboratory test report, when taking pictures by choose different angle, away from
From and illumination, with the test sample storehouse of abundant image to be identified.
S1-2: image slant correction: first detect the angle of inclination of image, then use according to the angle of inclination detected
Radiation conversion carries out rotation correction to image, i.e. can get satisfactory image.
S1-3: image denoising: in this detailed description of the invention, uses bilateral Filter bilateral filtering to carry out image
Denoising, in conjunction with pixel value similarity and a kind of compromise of spatial neighbor degree of image, considers grey similarity simultaneously, permissible
Retain edge well.
S1-4: image gray processing: the picture of collected by camera is 24 true coloured pictures, needs to be converted into gray-scale map.Ash is represented with g
Gray value after degreeization, R, G, B represent the red, green, blue component in true coloured picture, g=0.299R+0.587G+0.114B respectively.
S1-5: image binaryzation: character picture binaryzation is usually and the gray-scale map of 256 colors is converted into only black-and-white two color
Binary map.By selecting suitable threshold value T, each pixel in scanogram and threshold value T compare the most one by one
Relatively, i.e.Wherein (x, y) represents the gray value after gray proces to f, and (x y) represents f '
The pixel value of image after binaryzation, thr is threshold value.In this detailed description of the invention, carry out selected threshold by local threshold.
S2: template matching process:
S2-1: template image (hereinafter referred to as template) and the feature extraction of image to be identified (test sample): choose wherein
One template, by the character area of template in units of adjacent connected domain, carries out the demarcation of rectangle frame, obtains the feature of template
Set:
Wherein,Represent the lateral coordinates of jth feature square, longitudinal coordinate, horizontal picture in template T
Element length, longitudinal length in pixels, j={1,2,3 ..., n};N represents the number of characteristic rectangle in this template T.
In like manner, the sample to be tested after pretreatment is carried out feature extraction, obtains the characteristic set of test sample:
Wherein,Represent the lateral coordinates of ith feature rectangle in test sample S, longitudinal coordinate,
Pixels across length, longitudinal length in pixels, i={1,2,3 ..., m};M represents the number of characteristic rectangle in current test sample S.
After obtaining the set feature of template T and sample S, the two is carried out zero normalization and dimension normalization.Return
The characteristic set of the template T ' after one change is respectively as follows:
Wherein,Represent the template T ' middle ith feature rectangle after normalization lateral coordinates,
Longitudinal coordinate, pixels across length, longitudinal length in pixels, n represents the number of this template middle characteristic rectangle of T '.Obtain normalizing simultaneously
The characteristic set of the sample S ' after change is:
Wherein,Represent the sample S ' middle ith feature rectangle after normalization lateral coordinates,
Longitudinal coordinate, pixels across length, longitudinal length in pixels, m represents the number of the current middle characteristic rectangle of test sample S '.
After obtaining the set feature of template T ' and sample S ', by asking characteristic set similarity sim between T ' and S '
Whether judgment sample mates with template.In theory, if the value of sim is 1, then it represents that mate completely between sample with template.Ask
The formula of characteristic set similarity is:Wherein, αi∩βjRepresent that the i-th of current image to be identified is special
Levy the rectangular area overlapping area size of the jth characteristic rectangle of rectangle and current template image, αi∪βiRepresent the most to be identified
The rectangular area union size of the ith feature rectangle of image and the jth characteristic rectangle of current template image.
When there is overlap the rectangular area of ith feature rectangle and jth characteristic rectangle, αi∩βjAnd αi∪βiCalculating
Method is as follows:
When the rectangular area of ith feature rectangle and jth characteristic rectangle does not has overlap, αi∩βjAnd αi∪βiMeter
Calculation method is as follows:
αi∩βj=0
The traversal normalization of S2-2: anchor point and size: current just in the template of comparison in order to test sample is normalized to,
Then must obtain anchor point and best scale normalization ratio accurately.Anchor point i.e. needs the reference point of normalized zero,
Template and test sample need to choose identical rectangular characteristic as anchor point, just can carry out correct zero normalization.
According to priori, each rectangular characteristic of laboratory test report template is known, chooses current template here
One of them feature specified, as anchor point, then searches for the anchor point of correspondence in test sample.
See Fig. 2, different anchor points and different scalings and constitute a coordinate system.In a coordinate system, by different
Anchor point and different scalings travel through.When choosing different anchor points and scaling, the calculating according to sim is public respectively
Formula carries out characteristic of correspondence set Similarity Measure, obtains the characteristic set similarity set after all traversals, and therefrom chooses
The maximum, as the matching similarity of current template, is designated as SIM.
Before traversal anchor point, can first carry out Preliminary screening, reduce traversal scope.Known R is optimal scaling value
Big probable value, before traversal scaling, first scaling is fixed tentatively as R, then utilizes climbing algorithm to carry out screening and obtain
Optimal scaling (optimum shown in Fig. 2).Wherein R=ws/wTOr R=hs/hT, parameter wsImage for current test sample
Width, wTHorizontal picture traverse for current template;Parameter hsFor the picture traverse of current test sample, hTHorizontal stroke for current template
Picture traverse.
S2-3 template COLLECTION TRAVERSALSThe: current test sample is compared successively with 13 templates.When comparison, according to
Step in S2-2, by the anchor point of the anchor point of test sample and dimension normalization to template and yardstick, finally gives 13 coupling phases
Set like degree: { SIM1,SIM2,...,SIM13, take the template corresponding to the maximum in 13 matching similarities as working as
The Optimum Matching template of front test sample.
S3: Text region exports:
S3-1 image text positions: the space of a whole page feature of 13 laboratory test report templates is carried out typing, in laboratory test report template
Each project carries out String localization, method during location, each project and character area to be identified being used picture rectangle frame
Demarcate, each project is numbered simultaneously.According to the step in S2, determining the corresponding optimum mould of test sample
After plate, i.e. can get the space of a whole page feature of test sample.Now, the rectangle frame in template is positioned Information application to test sample
In, the locking of rectangle frame uses self-adapted search method to be accurately positioned.
S3-2 character classification identification: after image text positions successfully, knows carrying out classification according to projects different in template
Not.The result identified shows or passes through printer output result or the field according to system application by computer display
This FIELD Data is incorporated in system, obtains whole relevant informations of this character.
Claims (2)
1. the template matching method processed for Text region, it is characterised in that comprise the following steps:
Image to be identified and several template images are carried out Image semantic classification, described Image semantic classification include image slant correction,
Image denoising, image gray processing, image binaryzation;
Image to be identified and each template image are carried out feature extraction, obtain characteristic of correspondence set:
To the character area of each template image in units of adjacent connected domain, carry out the demarcation of rectangle frame, obtain template image
Characteristic setWhereinRepresent that the lateral coordinates of jth feature square in current template image T, longitudinal coordinate, pixels across are long
Degree, longitudinal length in pixels, j={1,2,3 ..., n}, n represent the number of characteristic rectangle in template image T;
To the character area of image to be identified in units of adjacent connected domain, carry out the demarcation of rectangle frame, obtain figure to be identified
The characteristic set of picture
WhereinRepresent the lateral coordinates of ith feature rectangle, longitudinal coordinate, horizontal stroke in current image S to be identified
To length in pixels, longitudinal length in pixels, i={1,2,3 ..., m}, m represent the number of characteristic rectangle in image S to be identified;
Template image and image to be identified carry out zero normalized, dimension normalization processes, and the most once returns
One change processes, then calculate a characteristic set similarity, takes the maximum characteristic set similarity coupling as current template image
Similarity;Take the Optimum Matching template as current image to be identified of the template image corresponding to maximum match similarity;
Wherein, the computing formula of characteristic set similarity sim is:Wherein, αi∩βjRepresent and currently treat
Identify the ith feature rectangle of image and the rectangular area overlapping area size of the jth characteristic rectangle of current template image, αi
∪βiRepresent the rectangular area of the ith feature rectangle of front image to be identified and the jth characteristic rectangle of current template image also
Collection size.
2. the method for claim 1, it is characterised in that when dimension normalization processes, first scaling is set to R
And carry out dimension normalization and process, then obtain optimal scaling based on climbing algorithm screening and carry out the yardstick of correspondence and return
One change processes, wherein R=ws/wT, parameter wsFor the picture traverse of current image to be identified, wTHorizontal figure for current template image
Image width degree;Or R=hs/hT, parameter hsFor the picture traverse of current image to be identified, hTHorizontal image for current template image
Width.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610284102.XA CN105913093B (en) | 2016-05-03 | 2016-05-03 | A kind of template matching method for Text region processing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610284102.XA CN105913093B (en) | 2016-05-03 | 2016-05-03 | A kind of template matching method for Text region processing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105913093A true CN105913093A (en) | 2016-08-31 |
CN105913093B CN105913093B (en) | 2019-06-21 |
Family
ID=56753487
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610284102.XA Expired - Fee Related CN105913093B (en) | 2016-05-03 | 2016-05-03 | A kind of template matching method for Text region processing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105913093B (en) |
Cited By (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105354566A (en) * | 2015-10-26 | 2016-02-24 | 江苏国光信息产业股份有限公司 | Character contamination removing method for prefix number identification |
CN106569059A (en) * | 2016-11-01 | 2017-04-19 | 广西电网有限责任公司电力科学研究院 | Production service system with function of converting test data to structured storage |
CN106708793A (en) * | 2016-12-06 | 2017-05-24 | 掌阅科技股份有限公司 | Annotation subscript recognition method, device and electronic equipment |
CN107220634A (en) * | 2017-06-20 | 2017-09-29 | 西安科技大学 | Based on the gesture identification method for improving D P algorithms and multi-template matching |
CN107403176A (en) * | 2017-03-29 | 2017-11-28 | 北京粉笔未来科技有限公司 | The detection method and device of character background inverse in textual image |
CN108230538A (en) * | 2017-12-30 | 2018-06-29 | 深圳怡化电脑股份有限公司 | A kind of bank note identification method, device, equipment and storage medium |
CN108961532A (en) * | 2017-05-26 | 2018-12-07 | 深圳怡化电脑股份有限公司 | Crown word number image processing method, device, equipment and storage medium |
CN108984578A (en) * | 2017-05-31 | 2018-12-11 | 株式会社日立制作所 | Computer, document recognition methods and system |
CN109086756A (en) * | 2018-06-15 | 2018-12-25 | 众安信息技术服务有限公司 | A kind of text detection analysis method, device and equipment based on deep neural network |
CN109271910A (en) * | 2018-09-04 | 2019-01-25 | 阿里巴巴集团控股有限公司 | A kind of Text region, character translation method and apparatus |
CN110569850A (en) * | 2019-08-20 | 2019-12-13 | 北京旷视科技有限公司 | character recognition template matching method and device and text recognition equipment |
CN110751682A (en) * | 2019-10-28 | 2020-02-04 | 普联技术有限公司 | Method, device, terminal equipment and storage medium for extracting and identifying image |
CN110910841A (en) * | 2019-12-16 | 2020-03-24 | 电子科技大学中山学院 | System and method for reducing ghost image of electrophoretic electronic paper |
CN110957016A (en) * | 2019-11-21 | 2020-04-03 | 山东鲁能软件技术有限公司 | Physical examination data intelligent recognition system and method based on health cloud management platform |
CN111046736A (en) * | 2019-11-14 | 2020-04-21 | 贝壳技术有限公司 | Method, device and storage medium for extracting text information |
CN111309618A (en) * | 2020-02-24 | 2020-06-19 | 广州市百果园信息技术有限公司 | Page element positioning method, page testing method and related device |
CN111421539A (en) * | 2020-04-01 | 2020-07-17 | 电子科技大学 | Industrial part intelligent identification and sorting system based on computer vision |
WO2020173008A1 (en) * | 2019-02-27 | 2020-09-03 | 平安科技(深圳)有限公司 | Text recognition method and apparatus |
CN112308046A (en) * | 2020-12-02 | 2021-02-02 | 龙马智芯(珠海横琴)科技有限公司 | Method, device, server and readable storage medium for positioning text region of image |
CN113313217A (en) * | 2021-07-31 | 2021-08-27 | 北京惠朗世纪科技有限公司 | Method and system for accurately identifying dip angle characters based on robust template |
CN114842003A (en) * | 2022-07-04 | 2022-08-02 | 杭州健培科技有限公司 | Medical image follow-up target pairing method, device and application |
CN117911732A (en) * | 2024-03-19 | 2024-04-19 | 中船黄埔文冲船舶有限公司 | Robot polishing rule template matching method and device |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102663377A (en) * | 2012-03-15 | 2012-09-12 | 华中科技大学 | Character recognition method based on template matching |
CN104182750A (en) * | 2014-07-14 | 2014-12-03 | 上海交通大学 | Extremum connected domain based Chinese character detection method in natural scene image |
CN105373794A (en) * | 2015-12-14 | 2016-03-02 | 河北工业大学 | Vehicle license plate recognition method |
-
2016
- 2016-05-03 CN CN201610284102.XA patent/CN105913093B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102663377A (en) * | 2012-03-15 | 2012-09-12 | 华中科技大学 | Character recognition method based on template matching |
CN104182750A (en) * | 2014-07-14 | 2014-12-03 | 上海交通大学 | Extremum connected domain based Chinese character detection method in natural scene image |
CN105373794A (en) * | 2015-12-14 | 2016-03-02 | 河北工业大学 | Vehicle license plate recognition method |
Non-Patent Citations (3)
Title |
---|
尹芳: "场景文本识别关键技术研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
李蕊: "中文印刷体文档数学公式识别***", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
王慧: "基于模板匹配的手写体字符识别算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (33)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105354566A (en) * | 2015-10-26 | 2016-02-24 | 江苏国光信息产业股份有限公司 | Character contamination removing method for prefix number identification |
CN105354566B (en) * | 2015-10-26 | 2019-01-22 | 江苏国光信息产业股份有限公司 | Character for crown word number identification pollutes minimizing technology |
CN106569059A (en) * | 2016-11-01 | 2017-04-19 | 广西电网有限责任公司电力科学研究院 | Production service system with function of converting test data to structured storage |
CN106708793A (en) * | 2016-12-06 | 2017-05-24 | 掌阅科技股份有限公司 | Annotation subscript recognition method, device and electronic equipment |
CN107403176B (en) * | 2017-03-29 | 2020-02-28 | 北京猿力未来科技有限公司 | Method and device for detecting reverse color of text background in text picture |
CN107403176A (en) * | 2017-03-29 | 2017-11-28 | 北京粉笔未来科技有限公司 | The detection method and device of character background inverse in textual image |
CN108961532A (en) * | 2017-05-26 | 2018-12-07 | 深圳怡化电脑股份有限公司 | Crown word number image processing method, device, equipment and storage medium |
CN108984578A (en) * | 2017-05-31 | 2018-12-11 | 株式会社日立制作所 | Computer, document recognition methods and system |
CN108984578B (en) * | 2017-05-31 | 2022-03-11 | 株式会社日立制作所 | Computer, document identification method and system |
CN107220634A (en) * | 2017-06-20 | 2017-09-29 | 西安科技大学 | Based on the gesture identification method for improving D P algorithms and multi-template matching |
CN108230538A (en) * | 2017-12-30 | 2018-06-29 | 深圳怡化电脑股份有限公司 | A kind of bank note identification method, device, equipment and storage medium |
CN108230538B (en) * | 2017-12-30 | 2020-05-15 | 深圳怡化电脑股份有限公司 | Paper money identification method, device, equipment and storage medium |
CN109086756A (en) * | 2018-06-15 | 2018-12-25 | 众安信息技术服务有限公司 | A kind of text detection analysis method, device and equipment based on deep neural network |
JP2020527260A (en) * | 2018-06-15 | 2020-09-03 | ジョンアン インフォメーション テクノロジー サービシズ カンパニー リミテッド | Text detection analysis methods, devices and devices |
WO2019238063A1 (en) * | 2018-06-15 | 2019-12-19 | 众安信息技术服务有限公司 | Text detection and analysis method and apparatus, and device |
CN109271910A (en) * | 2018-09-04 | 2019-01-25 | 阿里巴巴集团控股有限公司 | A kind of Text region, character translation method and apparatus |
WO2020173008A1 (en) * | 2019-02-27 | 2020-09-03 | 平安科技(深圳)有限公司 | Text recognition method and apparatus |
CN110569850B (en) * | 2019-08-20 | 2022-07-12 | 北京旷视科技有限公司 | Character recognition template matching method and device and text recognition equipment |
CN110569850A (en) * | 2019-08-20 | 2019-12-13 | 北京旷视科技有限公司 | character recognition template matching method and device and text recognition equipment |
CN110751682A (en) * | 2019-10-28 | 2020-02-04 | 普联技术有限公司 | Method, device, terminal equipment and storage medium for extracting and identifying image |
CN110751682B (en) * | 2019-10-28 | 2022-07-05 | 普联技术有限公司 | Method, device, terminal equipment and storage medium for extracting and identifying image |
CN111046736A (en) * | 2019-11-14 | 2020-04-21 | 贝壳技术有限公司 | Method, device and storage medium for extracting text information |
CN110957016A (en) * | 2019-11-21 | 2020-04-03 | 山东鲁能软件技术有限公司 | Physical examination data intelligent recognition system and method based on health cloud management platform |
CN110957016B (en) * | 2019-11-21 | 2023-08-08 | 山东鲁能软件技术有限公司 | Physical examination data intelligent identification system and method based on health cloud management platform |
CN110910841A (en) * | 2019-12-16 | 2020-03-24 | 电子科技大学中山学院 | System and method for reducing ghost image of electrophoretic electronic paper |
CN111309618A (en) * | 2020-02-24 | 2020-06-19 | 广州市百果园信息技术有限公司 | Page element positioning method, page testing method and related device |
CN111421539A (en) * | 2020-04-01 | 2020-07-17 | 电子科技大学 | Industrial part intelligent identification and sorting system based on computer vision |
CN112308046A (en) * | 2020-12-02 | 2021-02-02 | 龙马智芯(珠海横琴)科技有限公司 | Method, device, server and readable storage medium for positioning text region of image |
CN113313217A (en) * | 2021-07-31 | 2021-08-27 | 北京惠朗世纪科技有限公司 | Method and system for accurately identifying dip angle characters based on robust template |
CN113313217B (en) * | 2021-07-31 | 2021-11-02 | 北京惠朗世纪科技有限公司 | Method and system for accurately identifying dip angle characters based on robust template |
CN114842003A (en) * | 2022-07-04 | 2022-08-02 | 杭州健培科技有限公司 | Medical image follow-up target pairing method, device and application |
CN114842003B (en) * | 2022-07-04 | 2022-11-01 | 杭州健培科技有限公司 | Medical image follow-up target pairing method, device and application |
CN117911732A (en) * | 2024-03-19 | 2024-04-19 | 中船黄埔文冲船舶有限公司 | Robot polishing rule template matching method and device |
Also Published As
Publication number | Publication date |
---|---|
CN105913093B (en) | 2019-06-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105913093A (en) | Template matching method for character recognizing and processing | |
CN111612763B (en) | Mobile phone screen defect detection method, device and system, computer equipment and medium | |
CN106529537B (en) | A kind of digital instrument reading image-recognizing method | |
EP3309703B1 (en) | Method and system for decoding qr code based on weighted average grey method | |
CN104751187B (en) | Meter reading automatic distinguishing method for image | |
CN104751142B (en) | A kind of natural scene Method for text detection based on stroke feature | |
CN107491730A (en) | A kind of laboratory test report recognition methods based on image procossing | |
CN103439348B (en) | Remote controller key defect detection method based on difference image method | |
CN102704215B (en) | Automatic cutting method of embroidery cloth based on combination of DST file parsing and machine vision | |
CN104700072B (en) | Recognition methods based on lane line historical frames | |
CN104156693B (en) | A kind of action identification method based on the fusion of multi-modal sequence | |
CN104568986A (en) | Method for automatically detecting printing defects of remote controller panel based on SURF (Speed-Up Robust Feature) algorithm | |
CN111915704A (en) | Apple hierarchical identification method based on deep learning | |
CN104992449A (en) | Information identification and surface defect on-line detection method based on machine visual sense | |
CN110097596A (en) | A kind of object detection system based on opencv | |
CN115184380B (en) | Method for detecting abnormity of welding spots of printed circuit board based on machine vision | |
CN104597057A (en) | Columnar diode surface defect detection device based on machine vision | |
CN106096667A (en) | Bill images sorting technique based on SVM | |
CN104463138A (en) | Text positioning method and system based on visual structure attribute | |
CN106682678A (en) | Image angle point detection and classification method based on support domain | |
CN103699876B (en) | Method and device for identifying vehicle number based on linear array CCD (Charge Coupled Device) images | |
CN113657339A (en) | Instrument pointer counting and reading method and medium based on machine vision | |
CN103955925B (en) | The improvement probability Hough transformation curve detection method of minimum sampling is fixed based on piecemeal | |
CN109255794A (en) | A kind of full convolution edge feature detection method of standard component depth | |
CN105930813A (en) | Method for detecting line text under any natural scene |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190621 |